In the Long Term Evolution (LTE for short) system, cyclic shift sequences of Zadoff-Chu (ZC for short) sequences are used as the preamble by the Random Access Channel (RACH for short). These cyclic shift sequences are also referred to as Zero Correlation Zone (ZCZ for short) sequences.
In practical systems, after a mobile phone is powered on, downlink synchronization is first performed, and then the detection of the Broadcast Channel (BCH for short) is initiated. A base station informs, via the BCH channel, the mobile phone of the index and the step length of the cyclic shift of the first ZC sequence available for the RACH of the current cell. According to the index, the mobile phone makes use of certain mapping rule to calculate the serial number of the corresponding ZC sequence, and then, generates usable ZCZ sequences according to the step length of the cyclic shift and a certain “cyclic shift limitation rule” (the mobile phone is under a high speed circumstance).
If the number of the ZCZ sequences is smaller than a certain threshold Q, the mobile phone automatically increments the sequence index, and continuously generates the ZCZ sequences using the next ZC sequence, until the total number of the ZCZ sequences is larger than or equal to Q. Finally, the mobile phone randomly selects one sequence from all the generated usable ZCZ sequences as a preamble to be sent.
In the high speed circumstance, the frequency offset caused by Doppler Effect will generate, during the process of the preamble detection, a correlation peak alias, which will lead to a timing offset and a false detection. This problem is settled in the LTE system through limiting the use of some cyclic shifts according to a certain rule, which is the above mentioned “cyclic shift limitation rule”. Meanwhile, the cyclic shift limitation rule also limits the maximum cyclic shift NCS corresponding to each ZC sequence, and this maximum cyclic shift directly determines the maximum cell radius supported by each ZC sequence. Assuming that the distance between the correlation peak and the correlation peak alias thereof is du, the relation between the maximum cyclic shift NCS and du is:NCS=min(du,NZC−2·du)  (1)
Wherein, NZC is the length of a ZC sequence, du can be calculated by the following formula:
                    du        =                  {                                                                                                                                        m                        ·                                                  N                          ZC                                                                    -                      1                                        u                                    ,                                                                                                  when                    ⁢                                                                                  ⁢                                                                                            m                          ·                                                      N                            ZC                                                                          -                        1                                            u                                                        ≤                                      floor                    ⁡                                          (                                              N                        /                        2                                            )                                                                                                                                                                                      N                      ZC                                        -                                                                                            m                          ·                                                      N                            ZC                                                                          -                        1                                            u                                                        ,                                                                                                  when                    ⁢                                                                                  ⁢                                                                                            m                          ·                                                      N                            ZC                                                                          -                        1                                            u                                                        >                                      floor                    ⁡                                          (                                              N                        /                        2                                            )                                                                                                                              (        2        )            
Wherein, u is the serial number of the ZC sequence, and m is the minimum positive integer which makes
            m      ·              N        ZC              -    1    ua positive integer.
The mapping process between the indices and the serial numbers of the ZC sequences is actually the process of re-sequencing the ZC sequences. Wherein, the generation formula of the ZC sequences is shown as Equation (3) (0≦u≦NZC−1), and the serial number of the ZC sequence is the serial number used in the generation of each ZC sequence. The index of the sequence is the sequence number of each ZC sequence in a queue of sequenced ZC sequences, where the ZC sequences are sequenced according to a certain criterion.
                                                        x              u                        ⁡                          (              n              )                                =                      ⅇ                                          -                j                            ⁢                                                          ⁢                                                                                                        ⁢                                      π                    ⁢                                                                                  ⁢                                          un                      ⁡                                              (                                                  n                          +                          1                                                )                                                                                                              N                  ZC                                                                    ,                  0          ≤          n          ≤                                    N              ZC                        -            1                                              (        3        )            
At present, there are mainly two sequencing methods: one is to sequence according to the cubic metric (CM for short, it is a standard for measuring the Peak-to-Average Power Ratio of the emitted data, the larger the CM is, the higher the Peak-to-Average Power Ratio is) of the ZC sequences, and the other is to sequence according to the maximum cell radius supported by each ZC sequence under a high speed circumstance.
The first method is advantageous in that network planning of the ZC sequences can be conveniently performed according to the CM of a root sequence so as to assign the sequences with smaller CMs to the cells with larger radius, and the sequences with similar CMs to the same cell. Its shortcoming lies in that sequence fragments will be generated under the high speed circumstance, which will cause the waste of the sequences. In other words, during the process of generating the ZCZ sequences with the continuous incrementation of the sequence index, if the maximum cell radius supported by a ZC sequence is smaller than the radius of the current cell, this sequence neither could be used by the current cell, nor it could be used by other cells having radiuses smaller than the maximum cell radius supported by this ZC sequence (this is because that the index is continuously incremental, as shown in FIG. 1).
The second method is advantageous in avoiding the generation of the sequence fragments, that is disadvantageous in that the CMs of the ZC sequences assigned to a cell differs greatly from each other so that sequence planning can not be performed according to the CM.
A patent application of the present applicant with an application number of 200710135732.1 provides a method and device for sequencing the ZC sequences of the random access channel, which can not only prevent sequence fragments from being generated in the high speed circumstance, but also perform sequence planning and assigning according to the CM character of the sequences. The method comprises the following steps:
step 1, the sequences are sequenced according to the CM values of the sequences;
step 2, the sequencing result of step 1 is grouped according to a particular CM value (for example, the grouping may be performed according to the CM value of QPSK (about 1.2 dB));
step 3, the sequences in each group are sequenced, according to the maximum cell radius supported by the sequences under the high speed circumstance, for a second time to obtain the final sequencing result.
The present invention modifies the sequencing method and device provided by the patent application with the application number of 200710135732.1. The present invention not only inherits all the advantages of the application, but also performs a better sequence planning and assigning according to the CM character of the sequence, so that sequences with similar CMs are assigned to the same cell.